PURPOSES : This study aimed to investigate the factors affecting the severity of traffic crashes caused by personal mobility (PM) devices compared with those involving victims. METHODS : Traffic crashes involving PM devices were used to build a non-parametric statistical model using a classification tree. Based on the results, the factors influencing both at-fault and victim-related crashes caused by PM devices were analyzed. The factors affecting accident severity were also compared. RESULTS : Common factors affecting the severity of traffic crashes involving both perpetrators and victims using PM devices include occurrences at intersections, crosswalks at intersections, single roads, and inside tunnels. Traffic law violations by PM device users (perpetrators) influence the severity of crashes. Meanwhile, factors such as the behavior of perpetrators using other modes of transportation, rear-end collisions, road geometry, and weather conditions affect the severity of crashes where PM device users are the victims. CONCLUSIONS : To reduce the severity of traffic crashes involving PM devices, it is essential to extend the length of physically separated shared paths for cyclists and pedestrians, actively enforce laws to prevent violations by PM device users, and provide systematic and regular educational programs to ensure safe driving practices among PM device users.

PURPOSES : According to government data, the Black Spot Program has resulted in an average 28.8% reduction in traffic accidents within one year of project implementation in areas where road conditions improved. However, there has been a lack of in-depth analysis of the midto- long-term effects, with a predominant focus on short-term results. This study aimed to analyze the mid-to-long-term effects of the Black Spot Program to assess the sustainability of its reported short-term impact. Additionally, the differences in the mid-to-long-term effects were investigated based on the scale of traffic accidents at intersections and the characteristics of these effects are revealed. METHODS : The mid-to-long-term effects of the Black Spot Program were analyzed at 122 intersections in Seoul, Korea, where the program was implemented between 2013 and 2017, using traffic accident data spanning five years before and after implementation. Additionally, the differences in the program's effects were analyzed at the top-100 intersections with the highest traffic accident concentration in Seoul using the chi-square test to identify these differences. To theoretically validate these differences, the Hurst exponent, commonly used in economics, was applied to analyze the regression to the mean of the intersections and reveal the correlation with improvement. RESULTS : Through the Black Spot Program at 122 intersections, a 33.3% short-term accident reduction was observed. However, the midto- long-term effect analysis showed a 25.8% reduction, indicating a slightly smaller effect than previously reported. Specifically, the top-100 intersections exhibit a 15.4% reduction. A chi-square test with a 95% confidence level indicated significant differences in the program’s effects based on the scale of traffic accidents at intersections. The Hurst index (H ) was measured for the top-100 intersections, yielding H = 0.331. This is stronger than the overall H = 0.382 for all intersections in Seoul, suggesting that the regression to the mean is more pronounced, which may lead to a lower effectiveness of the improvement. CONCLUSIONS : The mid-to-long-term effects of the Black Spot Program were relatively lower than its short-term effects, with larger differences in effectiveness observed based on the scale of traffic accidents at intersections. This indicates the need to redefine the criteria for selecting project targets by focusing on intensive improvements at intersections, where significant effects can be achieved.

PURPOSES : This study aimed to derive the factors that contribute to crash severity in mixed traffic situations and suggest policy implications for enhancing traffic safety related to these contributing factors. METHODS : California autonomous vehicle (AV) accident reports and Google Maps based on accident location were used to identify potential accident severity-contributing factors. A decision tree analysis was adopted to derive the crash severity analyses. The 24 candidate variables that affected crash severity were used as the decision tree input variables, with the output being the crash severity categorized as high, medium, and low. RESULTS : The crash severity contributing factor results showed that the number of lanes, speed limit, bus stop, AV traveling straight, AV turning left, rightmost dedicated lane, and nighttime conditions are variables that affect crash severity. In particular, the speed limit was found to be a factor that caused serious crashes, suggesting that the AV driving speed is closely related to crash severity. Therefore, a speed management strategy for mixed traffic situations is proposed to decrease crash severity and enhance traffic safety. CONCLUSIONS : This paper presents policy implications for reducing accidents caused by autonomous and manual vehicle interactions in terms of engineering, education, enforcement, and governance. The findings of this study are expected to serve as a basis for preparing preventive measures against AV-related accidents.

2020년 국토교통부에서는 ‘결빙 취약구간 평가 세부 배점표’에 의하면, 전국의 고속국도 및 일반국도를 대상으로 결빙 취약 구간 464 개소를 선정하여 관리중에 있다. 그러나 감사원은 2020년 진행한 주요 사회기반시설(도로ㆍ고속철도) 안전관리실태 감사에서 결빙 취 약 구간 선정 시 터널 입출구부 등 결빙위험이 큰 구간이 도로포장 홈파기 대상구간에서 누락된 점을 지적하였다. 이러한 근거로 결 빙에 취약한 터널 입ㆍ출구에서 결빙사고가 우려되는 등 ‘겨울철 도로교통 안전 강화대책’의 실효성이 저하될 가능성이 제시되었다. 또한 본 연구에서 자체적으로 검토한 결과, 4개 특성 12개 항목으로 구성된 ‘결빙 취약구간 평가 세부 배점표’의 도로시설 항목에서 터널, 교량 등 도로시설물의 배점 부여 기준을 확인하기 어려웠으며, 각 도로시설에 대한 정의가 모호하여 평가표의 현장 적용성이 제 한되거나 신뢰도 검증이 부족한 점을 확인하였다. 본 연구에서는 국토교통부에서 제공하는 노드(Node) 및 링크(Link) 기반의 국내 도로망 GIS(Geographic Information System)데이터 에 결빙사고 데이터의 위치정보를 결합하여 고속국도 및 일반국도의 터널 및 교량 등을 포함하는 도로시설물 및 그 주변에서 발생한 결빙사고 이력을 자료화하였다. 최종적으로 도로시설물별 결빙사고 발생 비율 및 사고 심각도(사망자, 부상자 수)에 대한 분석을 통해 도로시설물의 결빙사고 상관 정도와 영향 범위를 파악하였다.

PURPOSES : This study empirically examines the determinants of traffic accidents by focusing on the transport culture index. METHODS : Two-stage least-squares estimation using an instrumental variable is used as the identification strategy. As the instrumental variable of the transport culture index, its past values, particularly before the outbreak of COVID-19 in 2018 are used. RESULTS : The empirical results, considering the potential endogeneity of the transport culture index, show that areas with higher values of the index are likely to have fewer traffic accident casualties. Local governments of regions with relatively frequent traffic accidents can run campaigns for residents to fasten their seatbelts, causing reverse causation. Ignoring this type of endogeneity underestimates the importance of the index as a key determinant of traffic accidents. CONCLUSIONS : Several traffic accidents occur in Korea, e.g., 203,130 accidents with 291,608 injuries and 5,392 deaths. As traffic accidents cause social costs, such as delays in traffic flow and damage to traffic facilities, public interventions are required to reduce them. However, the first step in public intervention is to accurately understand the relationship between the degree of damage in traffic accidents and the transport-related attributes of the areas where the accidents occurred. Although the transport culture index appears to be an appropriate indicator for predicting local traffic accidents, its limitations as a comprehensive index need to be addressed in the future.

국토교통부는 2020년 '결빙 취약구간 평가 세부 배점표’에 따라, 전국의 고속국도와 일반국도를 대상으로 410개 구간의 결빙 취약구 간을 선정하였다. 그러나, 2021년 감사원의 결빙 취약구간 지정 적정성 감사 결과에서 감사원은 현재 지정ㆍ관리 중인 결빙 취약구간 및 결빙 취약구간 평가 세부 배점표의 적정성에 문제를 제기하였다. 이에, 국토교통부는 결빙 취약구간을 재지정하여 발표하였으나 그 에 대한 평가 및 지정 적정성 검증이 아직 이루어지지 않았다. 본 연구에서는 결빙 취약구간과 결빙사고 데이터의 위치정보를 수집하여 GIS(Geographic Information System) 데이터로 구축하고 맵핑(Mapping)하여 결빙 취약구간 내 결빙사고이력을 확인함으로서 결빙 취약구간의 결빙사고 예측성능을 평가하였다. 또한, 각 결빙 사고 발생지점에서 도로시설, 교통, 선형구조, 환경인자 데이터를 수집하여 분석한다. 이를 통해 결빙사고와 각 인자 간의 상관성을 파 악하고, 그 결과에 따라 결빙 취약구간 평가 세부 배점표의 평가항목 및 각 항목별 배점을 수정하고 보완함으로써 결빙 취약구간의 신뢰성을 제고한다.

겨울철 국내 도로 결빙으로 인한 교통사고가 증가하는 추세를 보이고 있으며 2018년~2022년까지 총 4,609건의 결빙 교통사고가 발 생하였다. 결빙 교통사고의 치사율은 2.3으로 일반적인 교통사고와 비교하여 높은 치사율을 보이며 최근 5년(2018~2022)동안 결빙 교 통사고로 인하여 107명이 사망자와 7,728명의 부상자가 발생하였다. 현재 국토교통부에서 제시한 결빙 취약구간 평가기준표에 따라 결 빙 위험 구간을 지정하고 있으나, 해당 기준은 결빙의 주요 요인으로 고려되는 기상조건을 충분히 반영하지 못하고 있다. 도로 결빙은 노면온도가 0℃ 이하이며 노면에 수분이 공급될 때 형성되며 기온, 구름량, 풍속, 풍향, 상대습도, 강수량 등의 기상인자들이 복합적으 로 작용하여 결빙이 발생한다는 점을 고려하였을 때, 기상 특성은 도로 결빙의 주요 인자로 판단된다. 따라서 국내 결빙 취약구간 평 가기준의 개선이 필요하며 본 연구의 목적은 국내 결빙 교통사고 데이터를 분석하고 결빙이 형성되는 기상 조건을 구체화하는 것이다. 분석을 위한 데이터로 2018년~2022년까지 5년동안 발생한 결빙사고 사례와 기상청 방재기상관측소(AWS)에서 제공하는 기상 데이터 를 적용하였다. 이후, 박스도표, 확률밀도함수 등의 통계분석을 적용하여 결빙 형성 기상 조건을 구체화하였다. 이를 통하여 기존 결빙 취약구간 평가기준의 기상학적 개선 방향성을 제시할 수 있으며 더 나아가 도로 결빙 예측 로직 개발을 기대할 수 있다.

PURPOSES : This study aimed to identify factors affecting the duration of traffic incidents in tunnel sections, as accidents in tunnels tend to cause more congestion than those on main roads. Survival analysis and a Cox proportional hazards model were used to analyze the determinants of incident clearance times. METHODS : Tunnel traffic accidents were categorized into tunnel access sections versus inner tunnel sections according to the point of occurrence. The factors affecting duration were compared between main road and tunnel locations. The Cox model was applied to quantify the effects of various factors on incident duration time by location. RESULTS : Key factors influencing mainline incident duration included collision type, driver behavior and gender, number of vehicles involved, number of accidents, and post-collision vehicle status. In tunnels, the primary factors identified were collision type, driver behavior, single vs multi-vehicle involvement, and vehicles stopping in the tunnel after collisions. Incidents lasted longest when vehicles stopped at tunnel entrances and exits. In addition, we hypothesize that incident duration in tunnels is longer than in main roads due to the reduced space for vehicle handling. CONCLUSIONS : These results can inform the development of future incident management strategies and congestion mitigation for tunnels and underpasses. The Cox model provided new insights into the determinants of incident duration times in constrained tunnel environments compared to open main roads.

PURPOSES : This study aims to analyze the causes of pedestrian traffic accidents on community roads. METHODS : This study collected variables affecting pedestrian traffic accidents on community roads based on field surveys and analyzed them using negative binomial regression and zero-inflated negative binomial regression models. RESULTS : Model analysis results showed that the negative binomial regression model is more suitable than the zero-inflation negative binomial regression model. Additionally, the segment length (m), pedestrian volume (persons/15 min), traffic volume (numbers/15 min.), the extent of illegal parking, pedestrian-vehicle conflict ratio, and one-way traffic (one: residential, two: commercial) were found to influence pedestrian traffic accidents on community roads. Model fitness indicators, comparing actual values with predicted values, showed an MPB of 1.54, MAD of 2.57, and RMSE of 7.03. CONCLUSIONS : This study quantified the factors contributing to pedestrian traffic accidents on community roads by considering both static and dynamic elements. Instead of uniformly implementing measures, such as pedestrian priority zones and facility improvements on community roads, developing diverse strategies that consider various dynamic factors should be considered.

PURPOSES : This study investigates the factors affecting extra-long tunnel accidents by integrating data on tunnel geometry, traffic flow, and traffic accidents and derives the underlying implications to mitigate the severity of accidents. METHODS : Two processes centered on three key data points (tunnel geometry, traffic flow, and traffic accidents) were used in this study. The first is to analyze the spatial characteristics of extra-long tunnel traffic accidents and categorize them from multiple perspectives. The other was to investigate the factors affecting extra-long tunnel traffic accidents using the equivalent property-damage-only (EPDO) of individual accidents and the aforementioned data as the dependent and independent variables, respectively, by employing an ordered logistic regression model. RESULTS : Gyeonggi-do, Gyeongsangnam-do, and Gangwon-do are three metropolitan municipalities that have a significant number of extra-long tunnel accidents; Busan and Seoul have the most extra-long tunnel accidents, accounting for 23.2% (422 accidents) and 18.6% (339 accidents) of the 1,821 accidents that occurred from 2007 to 2020, respectively. In addition, approximately 70% of extra-long tunnel traffic accidents occurred along tunnels with lengths of less than 2 km, and Seoul and Busan accounted for over 60% of the top 20 extra-long tunnels with accidents. Most importantly, the Hwangryeong (down) tunnel in Busan experienced the most extra-long tunnel traffic accidents, with 77 accidents occurring during the same period. As a result of the ordered logistic regression modeling with EPDO and multiple independent variables, the significant factors affecting the severity of extra-long tunnel traffic accidents were determined to be road type (freeway, local route, and metropolitan city road), traffic flow (speed), accident time (year, summer, weekend, and afternoon), accident type (rear end), traffic law violations (safe distance violation and center line violation), and offending vehicles (van, sedan, and truck). CONCLUSIONS : Based on these results, the following measures and implications for mitigating the severity of extra-long tunnel traffic accidents must be considered: upgrading the emergency response level of all road types to that of freeways and actively promoting techniques for regulating high-speed vehicles approaching and traversing within extra-long tunnels are necessary. In addition, the emergency response and preparation system should be reinforced, particularly when the damage from extra-long tunnel traffic accidents is more serious, such as during the summer, weekends, and afternoons. Finally, traffic law violations such as safe distance and centerline violations in extra-long tunnels should be prohibited.

PURPOSES : This study empirically analyzes the determinants of fatal accidents based on raw data on traffic accidents occurring in Chungnam in 2020. METHODS : Regression models based on theoretical arguments for fatal traffic accidents are estimated using a binomial logit model. RESULTS : The prediction model for fatal accidents is affected by the degree of urbanization of the region, month and day of the accident, type of accident, and type of law violation. In addition, speeding or illegal U-turns among law violations appear more likely to result in fatal accidents. The road surface conditions at the time of the accident do not show a significant difference in the probability of fatality among traffic accidents. However, the probability of a fatal accident is rather lower in case of a snowy road; this is plausible, as drivers tend to drive more carefully in bad weather conditions. CONCLUSIONS : Among traffic accidents, fatal accidents appear to be affected by the time and place of the accident, type of accident, and weather conditions at the time of the accident. These analysis results suggest policy implications for reducing fatal accidents and can be used as a basis for establishing related policies.

해양사고 예방을 위해서는 사고의 원인과 결과에 대한 분석 및 진단뿐만 아니라, 사고의 발생 패턴과 변화 추이를 예측함으로 써 정량적 위험도를 제시할 필요성이 있다. 선박교통과 관련된 해양사고 예측은 선박의 충돌위험도 분석 및 항해 경로 탐색 등 선박교통 의 흐름에 관한 연구가 주로 수행되었으며, 해양사고의 발생 패턴에 대한 분석은 전통적인 통계 분석에 따라 제시되었다. 본 연구에서는 해양사고 통계 자료 중 선박교통관련 사고의 월별, 시간대별 발생 현황 데이터를 활용하여 해양사고 발생 예측 모델을 제시하고자 한다. 국내 해양사고 발생 현황 중 월별, 시간대별 데이터 집계가 가능한 1998년부터 2021년까지의 통계자료 중 선박교통 관련 데이터를 분류하 여 정형 시계열 데이터로 변환하였으며, 대표적인 인공지능 모델인 순환 신경망 기반 장단기 기억 신경망을 통하여 예측 모델을 구축하 였다. 검증데이터를 통하여 모델의 성능을 검증한 결과 RMSE는 초기 신경망 모델에서 월별 52.5471, 시간대별 126.5893으로 나타났으며, 관측값으로 신경망 모델을 업데이트한 결과 RMSE는 월별 31.3680, 시간대별 36.3967로 개선되었다. 본 연구에서 제안한 신경망 모델을 기 반으로 다양한 해양사고의 특징 데이터를 학습하여 해양사고 발생 패턴을 예측할 수 있을 것이다. 향후 해양사고 발생 위험의 정량적 제 시와 지역기반의 위험지도 개발 등에 관한 추가 연구가 필요하다.

PURPOSES : In this study, the factors affecting the severity of traffic accidents in highway tunnel sections were analyzed. The main lines of the highway and tunnel sections were compared, and factors affecting the severity of accidents were derived for each tunnel section, such as the tunnel access zone and tunnel inner zone. METHODS : An ordered probit model (OPM) was employed to estimate the factors affecting accident severity. The accident grade, which indicates the severity of highway traffic accidents, was set as the dependent variable. In addition, human, environmental, road condition, accident, and tunnel factors were collected and set as independent variables of the model. Marginal effects were examined to analyze how the derived influential factors affected the severity of each accident. RESULTS : As a result of the OPM analysis, accident factors were found to be influential in increasing the seriousness of the accident in all sections. Environmental factors, road conditions, and accident factors were identified as the main influential factors in the tunnel access zone. In contrast, accident and tunnel factors in the tunnel inner zone were found to be the influencing factors. In particular, it was found that serious accidents (A, B) occurred in all sections when a rollover accident occurred. CONCLUSIONS : This study confirmed that the influencing factors and the probability of accident occurrence differed between the tunnel access zone and inner zone. Most importantly, when the vehicle was overturned after the accident occurred, the results of the influencing factors were different. Therefore, the results can be used as a reference for establishing safety management strategies for tunnels or underground roads.

PURPOSES : In this study, we quantitatively prove the rubber necking phenomenon for highway traffic accidents and develop a calculation model based on the influencing factors. METHODS : Vehicle detector speed data in the opposite direction to the accident point were used based on the accident data on highways over the past three years, and a comparative verification was performed between nearby vehicle detector data to verify the reliability of the data. Accordingly, a binomial logistic model, ordinal probit regression model, and multilinear regression model were developed to compare the orientation. RESULTS : There was a difference in the influencing factors based on the dependent variable, and the day of the week, vehicle type, weather, longitudinal slope, and median height had an effect. Through a regression analysis, an influence coefficient was derived to calculate the driving speed deceleration value by rubbernecking. The results of the model analysis proved that the speed reduction caused by rubbernecking was more evident during the daytime than at night, during weekends compared to weekdays, and the speed reduction was more obvious for heavy vehicles compared to other types of vehicles. It can also be concluded that longer clearance time, higher accident severity, and higher traffic volume affect traffic delay. To verify the data and model equation, the mean prediction bias (MPB) and mean absolute deviation (MAD) were calculated for hundred cases randomly extracted from the collected accident data. These results were excellent. CONCLUSIONS : It can be developed into a human-engineered model that reflects various road/facility conditions, such as highways, other lanes, general roads, and roads without a median strip. This study is meaningful as a basic study on the quantitative effect of rubber necking.

PURPOSES : In this study, the installation of drowsy rest areas and accidents are analyzed. The factors that affected the accidents caused by drowsy drivers in rest areas are analyzed to improve the safety of rest areas. METHODS : By comparing and analyzing the installation status of the rest areas for drowsy drivers, the accident status were analyzed. The logistic regression model was used to analyze the factors that affect accidents in the drowsy rest area. RESULTS : Most rest areas were installed below the installation criteria. Several accidents occurred when the vehicle entered the drowsy rest area. These rest areas had a short entry ramp, and no safety facilities were installed. The logistic regression model showed that the risk of an accident is lowered when the deceleration lane is longer than 215 m. Additionally, the risk of an accident is lowered when the rest area is installed in the straight section or the curve section, wherein the curve radius is greater than 2 km. CONCLUSIONS : In this study, we evaluated the installation status of the rest areas for drowsy drivers by comparing installation elements. Most rest areas for drowsy drivers were installed at different lengths of the ramp. Some of these were installed on the slope or curved sections of the road. We analyzed the accident status and developed an accident modal using the logistic regression model to identify the factors that affect accidents. It will be necessary to analyze accidents in drowsy rest areas continuously to improve safety for drowsy drivers.

PURPOSES: The purpose of this study is to develop a traffic accident prediction model using statistical data, to analyze child traffic accidents in school zones. Furthermore, we analyze the factors affecting child traffic accidents, as obtained from the results of the developed model. METHODS : From the literature review, we obtained data for child traffic accidents and various variables relating to road geometry and traffic safety facilities in school zones. We used these variables and data to develop a child traffic accident analysis model. The model was then developed into three types using the Limdep 9.0 statistical tool. RESULTS: As a result of the overdispersion test, the Poisson regression model was applied to all types of models with an overdispersion coefficient of close to zero. The results of the model development are as follows. The main model (all scope of analysis) was for a kindergarten, considering a local roadway, the accessibility of the roadway, the number of unsignalized intersections, and the school zones in commercial area as variables that increase traffic accidents. Sub-model typeⅠ(only the roadway connected to the main entrance) was for a kindergarten, considering a local roadway, skid resistant pavement, no-parking signs, the number of unsignalized intersections, and the number of commercial facilities as variables that increase traffic accidents. The main model and sub-model type Ⅰ showed a reduction in accidents when using forward-type traffic signals. Sub-model typeⅡ(only the roadway not connected to the main entrance) shows that the local roadway is the variable that most increases the probability of traffic accidents. However, when the roadway and walkway are separated, the probability of traffic accidents decreases significantly, by up to 90%. CONCLUSIONS: The results of this study demonstrate the need to restructure the method used to improve school zones. Moreover, the effect of various traffic safety facilities was quantitatively analyzed.